Image processing method, image processing device and image processing system
An image processing method includes steps of receiving an image sequence; when at least one object appears in the image sequence, analyzing a moving trajectory of each object; extracting at least one characteristic point from each moving trajectory; classifying the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster; and storing at least one characteristic parameter of each cluster.
Latest VIVOTEK INC. Patents:
- IMAGE ANALYSIS MODEL ADJUSTMENT METHOD AND IMAGE ANALYSIS APPARATUS
- IMAGE ANALYSIS METHOD AND CAMERA APPARATUS
- IMAGE INFORMATION TRANSFORMATION METHOD, IMAGE INFORMATION OUTPUTTING METHOD, IMAGE ANALYSIS DEVICE
- IMAGE DEFECT IDENTIFICATION METHOD AND IMAGE ANALYSIS DEVICE
- Image calibration method and image calibration device
The invention relates to an image processing method, an image processing device, and an image processing system and, more particularly, to an image processing method, an image processing device, and an image processing system capable of reducing data storage amount effectively.
2. Description of the Prior ArtHeat map is used to represent the distribution and density of data points on a map or an image. To generate a heat map, the prior art determines the variations in continuous images based on the difference between images, so as to determine colors of the heat map, and stores the heat map by bitmap. Furthermore, another prior art stores moving trajectories of all objects in the images and then uses the stored data to draw the moving trajectories of all objects on the heat map. However, the data storage amount of the aforesaid manner is huge. The more the number of objects in the image is, the huger the data storage amount is. Consequently, hardware loading increases.
SUMMARY OF THE INVENTIONAn objective of the invention is to provide an image processing method, an image processing device, and an image processing system capable of reducing data storage amount effectively, so as to solve the aforesaid problems.
According to an embodiment of the invention, an image processing method comprises steps of receiving an image sequence; when at least one object appears in the image sequence, analyzing a moving trajectory of each object; extracting at least one characteristic point from each moving trajectory; classifying the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster; and storing at least one characteristic parameter of each cluster.
According to another embodiment of the invention, an image processing device comprises an image receiving unit, a storage unit, and a processing unit. The image receiving unit receives an image sequence. The processing unit is electrically connected to the image receiving unit and the storage unit. When at least one object appears in the image sequence, the processing unit analyzes a moving trajectory of each object, extracts at least one characteristic point from each moving trajectory, classifies the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster, and stores at least one characteristic parameter of each cluster into the storage unit.
According to another embodiment of the invention, an image processing system comprises a host device and an image processing device. The host device comprises a storage unit and a processing unit. The processing unit is electrically connected to the storage unit. The image processing device communicates with the host device. The image processing device comprises an image receiving unit. The image receiving unit receives an image sequence and transmits the image sequence to the host device. When at least one object appears in the image sequence, the processing unit analyzes a moving trajectory of each object, extracts at least one characteristic point from each moving trajectory, classifies the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster, and stores at least one characteristic parameter of each cluster into the storage unit.
As mentioned in the above, after receiving the image sequence, the invention analyzes the moving trajectory of each object in the image sequence, extracts the characteristic point from each moving trajectory, and classifies the characteristic point of each moving trajectory within the predetermined time period (e.g. 15 minutes, 30 minutes, etc.) into at least one cluster. Then, the invention stores the characteristic parameter of each cluster. The characteristic parameter of each cluster may be used to generate an image distribution map (e.g. heat map). Since the invention needs not to store the moving trajectory of each object, the invention can reduce data storage amount effectively, so as to reduce hardware loading.
These and other objectives of the present invention will no doubt become obvious to those of ordinary skill in the art after reading the following detailed description of the preferred embodiment that is illustrated in the various figures and drawings.
Referring to
As shown in
In this embodiment, the host device 10 comprises a storage unit 100 and a processing unit 102, wherein the processing unit 102 is electrically connected to the storage unit 100. In practical applications, the storage unit 100 may be a memory, a hard disc, or other data storage devices; and the processing unit 102 may be a processor or a controller with data processing function. In general, the host device 10 may be further equipped with some necessary hardware or software components for specific purposes, such as a circuit board, a power supply, applications, a communication module, etc., and it depends on practical applications.
In this embodiment, the image processing device 12 comprises an image receiving unit 120. In practical applications, the image receiving unit 120 may be a charge coupled device (CCD) sensor or a complementary metal-oxide-semiconductor (CMOS) sensor. In general, the image processing device 12 may be further equipped with some necessary hardware or software components for specific purposes, such as a circuit board, a power supply, applications, a communication module, a lens, etc., and it depends on practical applications.
The image processing system 1 of the invention can be used to generate an image distribution map, wherein the image distribution map may be, but not limited to, a heat map. When the image processing system 1 is used to generate the image distribution map, first of all, the image receiving unit 120 receives an image sequence 3 (step S10 in
When at least one object 30 appears in the image sequence 3, the processing unit 102 analyzes a moving trajectory of each object 30 (step S12 in
Then, the processing unit 102 classifies the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster (step S16 in
Then, the processing unit 102 stores at least one characteristic parameter of each cluster R1-R9 and D1-D4 into the storage unit 100 (step S18 in
Still further, the aforesaid characteristic parameter may further comprise a number of characteristic points in each cluster R1-R9 and D1-D4. For each cluster R1-R9, the number of characteristic points is equal to the number of begin points, end points, and/or turn points; and for each cluster D1-D4, the number of characteristic points is equal to the number of stop points.
Moreover, when the at least one characteristic point comprises a stop point, the aforesaid characteristic parameter may further comprise an averaged stop time of the stop point.
In this embodiment, the processing unit 102 may further analyze a number of moving times between every two clusters. For example, for the clusters R1 and R2, the processing unit 102 may accumulate a number of moving times when each object 30 moves from the cluster R1 to the cluster R2 and accumulate a number of moving times when each object 30 moves from the cluster R2 to the cluster R1. Then, the processing unit 102 may store the number of moving times between every two clusters into the storage unit 100 and generate an image distribution map according to the stored characteristic parameter and/or the number of moving times between every two clusters. In this embodiment, the image distribution map may be, but not limited to, a heat map.
Referring to
In addition, except to classify all of the characteristic points within the predetermined time period (e.g. 15 minutes) into a plurality of clusters, the invention may also combine all of characteristic points within a plurality of time periods (e.g. combine four continuous “15 minutes” time periods to obtain one “1 hour” time period or combine 2880 continuous “15 minutes” time periods to obtain one “1 month” time period) and then process all of the characteristic points within the combined time period by the aforesaid manner (e.g. draw the heat map and so on), so as to represent the statistic result within different time periods.
Furthermore, when analyzing the moving trajectory of each object 30, the invention may calculate a life cycle of each trajectory, i.e. the time period from the beginning to the end of each trajectory. Then, the invention may find out a trajectory with a life cycle less than a predefined shortest threshold (e.g. 3 minutes). For example, when the invention is applied to monitor and analyze retail images, the aforesaid data may represent that the stop time of a customer is abnormal (too short). Then, the invention may use the aforesaid manner to find out the characteristic points of the trajectory, classify the characteristic points into a plurality of clusters, and store the clusters into the database, so as to generate the heat map. Alternatively, the invention may find out a trajectory with a life cycle more than a predefined largest threshold (e.g. 10 minutes). For example, when the invention is applied to monitor and analyze retail images, the aforesaid data may represent that a customer is highly interested in a product. Then, the invention may use the aforesaid manner to find out the characteristic points of the trajectory, classify the characteristic points into a plurality of clusters, and store the clusters into the database, so as to generate the heat map.
It should be noted that the control logic of the image processing method of the invention can be implemented by software. Needless to say, each part or function of the control logic may be implemented by software, hardware or the combination thereof.
Referring to
In other words, the image processing method of the invention may be performed by the image processing device 12 with the host device 10 or performed by the image processing device 12′ according to practical applications.
As mentioned in the above, after receiving the image sequence, the invention analyzes the moving trajectory of each object in the image sequence, extracts the characteristic point from each moving trajectory, and classifies the characteristic point of each moving trajectory within the predetermined time period (e.g. 15 minutes, 30 minutes, etc.) into at least one cluster. Then, the invention stores the characteristic parameter of each cluster. The characteristic parameter of each cluster may be used to generate an image distribution map (e.g. heat map). Since the invention needs not to store the moving trajectory of each object, the invention can reduce data storage amount effectively, so as to reduce hardware loading.
Those skilled in the art will readily observe that numerous modifications and alterations of the device and method may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.
Claims
1. An image processing system comprising:
- a host device comprising a storage and a processor, the processor being electrically connected to the storage; and
- an image processing device communicating with the host device, the image processing device comprising an image receiver, the image receiver receiving an image sequence and transmitting the image sequence to the host device;
- wherein when at least one object appears in the image sequence, the processor analyzes a moving trajectory of each object, extracts at least one characteristic point from each moving trajectory, classifies the at least one characteristic point of each moving trajectory within a predetermined time period into at least one cluster, and stores at least one characteristic parameter of each cluster into the storage.
2. The image processing system of claim 1, wherein the at least one characteristic point comprises at least one of a begin point, an end point, a turn point, and a stop point.
3. The image processing system of claim 1, wherein each cluster is represented by a circle and the at least one characteristic parameter comprises a center and a radius of the circle.
4. The image processing system of claim 1, wherein the at least one characteristic parameter comprises a number of characteristic points in the cluster.
5. The image processing system of claim 1, wherein when the at least one characteristic point comprises a stop point, the at least one characteristic parameter comprises an averaged stop time of the stop point.
5163124 | November 10, 1992 | Yabe |
5764283 | June 9, 1998 | Pingali |
6520864 | February 18, 2003 | Wilk |
6526156 | February 25, 2003 | Black |
6778171 | August 17, 2004 | Kikinis |
7636093 | December 22, 2009 | Kuwamoto |
9245196 | January 26, 2016 | Marks |
20060045354 | March 2, 2006 | Hanna |
20060182316 | August 17, 2006 | Bang |
20120020518 | January 26, 2012 | Taguchi |
20130120405 | May 16, 2013 | Maloney |
20130268256 | October 10, 2013 | Dariush |
20140003741 | January 2, 2014 | Cui |
20140100828 | April 10, 2014 | Dariush |
20150094991 | April 2, 2015 | John |
20160076908 | March 17, 2016 | Pang |
20160148392 | May 26, 2016 | Hepper |
201033908 | September 2010 | TW |
Type: Grant
Filed: Sep 4, 2017
Date of Patent: Mar 17, 2020
Patent Publication Number: 20180096225
Assignee: VIVOTEK INC. (New Taipei)
Inventors: Cheng-Chieh Liu (New Taipei), Chih-Yen Lin (New Taipei)
Primary Examiner: Samir A Ahmed
Application Number: 15/694,997
International Classification: G06K 9/62 (20060101); G06K 9/46 (20060101); G06T 1/60 (20060101); G06T 11/20 (20060101); G06K 9/00 (20060101);